Bug 10581

Summary: Really really really slow graphics on intel 945 / acer laptop
Product: xorg Reporter: Robert Atwood <r.atwood>
Component: Driver/intelAssignee: Eric Anholt <eric>
Status: RESOLVED DUPLICATE QA Contact: Xorg Project Team <xorg-team>
Severity: major    
Priority: medium    
Version: git   
Hardware: x86 (IA32)   
OS: Linux (All)   
URL: http://www.geocities.com/robertcarlatwood/
Whiteboard:
i915 platform: i915 features:

Description Robert Atwood 2007-04-09 12:50:40 UTC 
This may be the same as bug 10542 but I am not knowledgeable enough to be certain. I get ridiculously slow performace from certain applications using OPENGL features. I am attaching my 'processing' sketch files, that run at video frame rates on my old (4-year old used) laptop (now dead so I cannot view log outputs or anything) 
On my new Acer Aspire 5612 laptop which has the Intel 945 builtin, using Debian Experimental for the x.org and drivers (1.9.94) and Mesa that I build from GIT myself, it runs at about 1 frame per second. To try this you would neet to install Processing, from http://processing.org , just put the whole tarball contents in /usr/local/processing and add to the path, to duplicate my installation, then run my 'sketch' in it's environment

Also Simon Yuell's Python-opengl example spinning cube (thus avoiding the Java layer of Processing) runs absurdly slow even if the 100 millisecond sleep  statement is removed.

I am also attaching my log files and output of xrandr --verbose as requested for bug 10542, in hopes that it will be helpful in resolving this matter. See the url 
http://www.geocities.com/robertcarlatwood/ 

I am afraid I don't have enough knowlege about how opengl really works to know how the 'processing' framework actually makes use of the opengl though. I'm looking for how to test the same feature outside this (Java-based) framework.


I am not sure which logs you wanted, since the message just said 'attach your logs and kernel output', so I am just putting the entire /var/log in a tarball. Pleas specifiy if this does not include what you need.



Thanks, 
Robert
Comment 1 Michel Dänzer 2007-04-10 00:40:02 UTC 
(In reply to comment #0)
> This may be the same as bug 10542 but I am not knowledgeable enough to be
> certain. 

It seems likely, as the characteristic feature of that bug seems to be that the X server starts up on display pipe B only, which is the case for you as well.

As mentioned in that bug, can you try whether the 'Possible fix' patch attached to bug 10546 helps? Also, going back to i915 DRM 1.5 or older should avoid this problem, e.g. using the Debian packaged 2.6.18 kernel.

> I am also attaching my log files and output of xrandr --verbose as requested
> for bug 10542, in hopes that it will be helpful in resolving this matter. 

It would only be really useful with xrandr 1.2 or newer.

> See the url http://www.geocities.com/robertcarlatwood/ 

In the future, please attach files to bug reports directly instead of putting them on external websites.

> I am not sure which logs you wanted, since the message just said 'attach your
> logs and kernel output', so I am just putting the entire /var/log in a tarball.

'log' in the X context usually refers to /var/log/Xorg.0.log, and the kernel output is shown by dmesg or in /var/log/kern.log (on Debian based systems).
Comment 2 Michel Dänzer 2007-04-10 02:28:28 UTC 
Assuming these are the same for now, we can always decouple them again if this turns out not to be the case.

*** This bug has been marked as a duplicate of bug 10542 ***

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.